Meta-analysis links GLP-1 drugs to retained stomach contents before endoscopy

People taking GLP-1 receptor agonists had roughly 4.5 times the odds of having food or fluid left in the stomach during an upper endoscopy compared with people not taking the drugs, according to a systematic review and meta-analysis published in the April 2025 issue of Gastrointestinal Endoscopy ^[1]. The same analysis found the same elevated odds for procedures that had to be stopped early.

The review pooled 23 observational studies with a total of 262,018 patients ^[1]. Researchers searched PubMed, Scopus, Web of Science and Cochrane Central from their start through July 1, 2024, and EMBASE from 1974 through June 28, 2024 ^[1]. The search turned up 3,671 records, 1,851 after duplicates were removed; reviewers screened 38 full texts and included 23 studies ^[1].

The numbers

GLP-1 receptor agonist users had a statistically significant increase in residual gastric contents, with an odds ratio of 4.54 (95% confidence interval 3.30–6.24; P < .00001) ^[1]. Premature endoscopy termination showed an odds ratio of 4.54 as well (95% CI 3.05–6.75; P < .00001) ^[1].

Aspiration pneumonia was different. The pooled odds ratio was 0.96 (95% CI 0.53–1.75; P = .90), which the authors describe as no significant difference between users and non-users ^[1].

One finding pointed the other way. When an upper endoscopy and a colonoscopy were done on the same day, rather than an upper endoscopy alone, the odds of residual gastric contents fell (odds ratio 0.28; 95% CI 0.22–0.36; P < .00001) ^[1].

The authors note that many of the included patients were following standard fasting rules — no solid food for at least eight hours, per 2017 American Society of Anesthesiologists practice guidelines — and still had retained contents ^[1]. They connect the finding to how these drugs work: GLP-1 receptor agonists slow stomach emptying as part of their intended effect ^[1].

What the analysis does not show

All 23 studies were observational, not randomized trials ^[1]. The pooled result for residual gastric contents had a heterogeneity score (I²) of 68 percent, which indicates the studies did not all point the same way ^[1].

The analysis does not report results broken out by individual drug, such as semaglutide (Ozempic, Wegovy, Rybelsus) or tirzepatide (Mounjaro, Zepbound), by dose, or by how long patients paused their medication before the procedure ^[1]. Whether holding a dose changes the risk has not been established by this analysis. One of the 14 authors disclosed a consulting relationship with Eli Lilly; the rest reported no financial relationships ^[1].

Why it matters for patients

Upper endoscopy is a common procedure, and it is usually done with sedation. Retained stomach contents raise the concern that food or liquid could enter the airway during sedation, which is why anesthesia and gastroenterology groups have been debating how to handle GLP-1 patients.

On the outcome patients worry about most — aspiration pneumonia — this analysis found no significant difference between GLP-1 users and non-users ^[1]. What it did find is that procedures were more likely to be cut short among users, which can mean a repeat procedure, another round of prep and fasting, and another day of lost time ^[1].

Telling the endoscopy team about any GLP-1 medication before the procedure is the practical step this evidence supports. Decisions about pausing or continuing a medication before a procedure are made by the prescribing clinician and the endoscopy or anesthesia team, and this analysis does not offer a schedule for doing so.

What happens next

No new professional guidance accompanied this analysis, and no update to the fasting instructions cited in it has been reported. The paper cites existing American Society of Anesthesiologists consensus guidance and a multisociety statement from several liver and gastroenterology organizations ^[1]. Whether future guidance changes for GLP-1 users undergoing endoscopy has not been disclosed.